Your browser (Internet Explorer 7 or lower) is out of date. It has known security flaws and may not display all features of this and other websites. Learn how to update your browser.


Publish or Perish…or at least forfeit your PhD

A couple of days ago I received an email addressed to the graduate student population in my department informing us that there would be a meeting in a couple of weeks to discuss implementing a publication requirement to receive a PhD. The email did not give the specifics, but it seems likely that the requirement will be to have published at least one first author paper in a peer-reviewed journal before graduating.

Let me give you a little more information about my department. It’s called MCB, for Molecular Biology, Cell Biology, and Biochemistry. It encompasses approximately 30 research groups primarily focused on primary research using various different model organisms. For example, I work on the process of RNA localization and concurrent translational repression in Xenopus laevis (African clawed frogs). The lab upstairs uses Drosophila melanogaster (fruit flies) to study tissue patterning. And downstairs there are three labs using Arabidopsis thaliana (a type of plant related to mustard) to study a variety of genetic and cell biological questions. We also have a fairly large contingent of bioinformaticists and structural biologists. So that should give you some idea of the variety of research contained within our department.

The graduate program is not massive (perhaps 50 students at any one time) but is fairly evenly distributed throughout these diverse labs (after completing a first year core course and three lab rotations). As you can imagine, therefore, each student has a very different lab experience, where different techniques are learned and different data analyzed. They also receive vastly different degrees of advice and mentorship depending on their group leader or P.I. (principle investigator) and the other members of the lab, such as postdocs, technicians, other grad students, and undergraduates.

At the end of our second year, we prepare a written research proposal that is then evaluated in a three-hour oral examination. This comprises the qualifying exam, and once this is passed, a student becomes a PhD candidate. They are deemed ready to undertake research into their chosen field.

I should also point out that we are lucky enough to be guaranteed funding throughout the course of our degree. There is no time limit imposed, although after five years it is expected that you are at least thinking of defending your thesis in the near future.

The thesis itself is usually made up of four or five chapters. The first is a comprehensive review of the literature and state of the field, and in many cases is published as such after the student defends. The remaining chapters summarize the student’s research, and are usually written in a similar format to a published, peer-reviewed article. In fact, if you have published papers, our department is quite happy for you to simply include the paper without re-writing it.

However, until now, there has been no requirement that any of the thesis chapters have to be published in the peer-reviewed literature. They must simply reflect significant insight into the field of study gained through well-controlled and well-executed experiments.

And isn’t that how it should remain?

I think so, but I thought it would be a good idea to increase my sample size, so I put the word out on Twitter. Sure enough my fabulous tweeps took up the debate, and I have storified a select sample here. Roughly the responses can be split into two camps:

Having a publication requirement is a good thing.

A number of programs, it turns out, already have in place a publication requirement. For some it is submission rather than acceptance, which seems a little more reasonable to me. Some see this requirement as a means of judging the value of a student’s thesis work:

Others see publication requirements as a challenging goal to work towards:

Of course publishing papers while in graduate school is incredibly valuable. Not only do you prepare your work in the most polished way possible and deal with the peer review system first hand, you make your transition up the academic ladder far more straightforward. In some fields it is impossible to get a postdoctoral position without a good publication record, and forget trying to get a faculty position without them. By requiring at least one first author paper, then, a graduate program is helping its students.

The major flaw here is that students rarely control the publication process. Step one is selecting a publishable set of experiments from the outset of your thesis. At least in my program, a committee made up of your adviser and three to four other faculty members assesses this in the qualifying exam. They consider whether the experiments are pertinent to the field and can be accomplished within a reasonable time frame. However, as anyone who has ever worked in research knows, an experiment on paper is a completely beast in the lab. Even with an intelligent hypothesis and a good set of hands a scientist can find themselves at the mercy of a dodgy water supply, or a cell line that just won’t grow, or a mutant that does not behaved as advertised in that Cell paper everyone loved. Even if a project doesn’t seem risky to a committee, it can turn risky pretty quickly. Enter the backup project. We all have them. In my case it has turned into my main project, and it has turned out to be just as risky. Awesomesauce.

Regardless of any issues in the lab, once you have a “publishable” set of data, the next step is to convince your P.I. that you should send it out for review. Once you get past that hurdle, you wait with baited breath while an editor at your journal-of-choice decides whether to send it out for review. If it gets reviewed you can virtually guarantee it will come back with comments that require further experiments to be performed. Even if you address all the comments, you are still at the mercy of a second round of review. Then your manuscript will finally see the light of day. If you don’t make it to review, you have to rewrite the paper into the format of your second choice journal and start the process all over again. I have known work sit in rounds of review for close to two years. While the topic of peer-review would in itself make for a very substantial blog post, my point here is merely that a prospective PhD student could end up mired in this mess for a really long time through no fault of their own.

As for the argument that having a published paper helps in the job hunt, there is no question that this is undoubtedly true. However, I know of several very talented scientists who got their postdoctoral positions in various labs at some of the most prestigious universities in the country (even the world) without a single peer-reviewed publication. What they did have were very strong letters of recommendation and presented themselves and their work well at interview. It should also be noted that pursuing an academic career is not the goal of every newly minted PhD, and there are other skills they possess from their training that should be valued equally with a publication record. (I went into this in detail in a previous post, but briefly I am referring to primarily to leadership and communications skills.)

What the anti-publication requirement folks had to say

A major concern voiced during this conversation was that putting the focus on what is “publishable” would keep students from tackling challenging and risky projects:

That negative data is not publishable but nonetheless represents the result of hard work, critical thinking, and good science:

And that there is considerable inequality between P.I.s and committees in terms of what constitutes an acceptable thesis:

“A candidate must submit a project or thesis or dissertation often consisting of a body of original academic research, which is in principle worthy of publication in a peer-refereed context.” (Wikipedia via Dinham and Scott)

Imagine if you were scooped, weeks before submission? If a lab at another institution managed to publish before you? Does that automatically mean that everything you did is no longer worthy of garnering you those three little letters after your name?

In conclusion…

The PhD experience is about so much more than publishing a paper (or two, or three…), and while navigating peer-review successfully is an important part of a grad student’s training, all the other aspects of this training should not be overlooked. Participation in conferences, outreach and teaching experience, written and verbal communication, and fruitful internal and external collaboration, are all crucial components of the degree. It would seem to me that while having a cut-and-dry system for figuring out when a student is ready to defend his or her thesis is a great idea, setting a publication requirement is not the way to do it. And if such a system is put in place, the faculty will be the ones responsible, as they are our mentors. They guide the project. They train us. They submit the papers. We are here to learn to become independent researchers and educators, not slaves to the peer-review process.

I’d love to hear your thoughts on this topic, so please, get your comment on!

2.22.12 UPDATE: I spoke to the director of the graduate program last week. After considering graduate students’ comments, the new rule will be that a student have at least one first author, or more than one other author, paper submitted before they defend. This decision is aimed at reducing the number of students that are “forced out” without a publication. I also raised with her the idea of adding value to other aspects of the PhD program, such as outreach, teaching, participation in conferences, and garnering independent funding. She agreed that these are important things to consider, but at this time there is nothing in the works to officially take them into account. She was, however, very receptive to the idea of setting up workshops covering topics relating to non-tenure-track career paths. These workshops would cover things like building an online presence by effectively using social media, networking, and other transferable skills relevant to life outside the academy.

While I am still dubious about the amount of emphasis placed on publication, I am happy that this compromise will not place students at the whim of the third reviewer. And, with an improved attitude towards leaving “traditional” academic career paths, hopefully the graduate program here is on the road to graduating better-prepared, and published, scientists.




The publication requirement seems like a terrible idea to me, for all the reasons that you discuss here. As you say, whether or not a bit of work is publishable can have very little to do with whether it represents high-quality science. Surely the latter is what the PhD examiners should be concerned with, and they ought to be able to evaluate that without farming it out for peer review, which can (as you say) be enormously time-consuming and inefficient.


You bring up another good point…published papers don’t always contain good science.


Absolutely. And if there is enough pressure to publish, doing good, thorough, honest science can even be an obstacle. Just ask Hauser and Stapel.


Ha! Excellent point :)

Zen Faulkes

This is one of these cases where a lot depends on how the policy is implemented. I think it is a perfectly reasonable thing to expect work from a doctorate to be published. But there are going to be exceptions.

Now, if the people running this will look at those cases, and will consider waiving the publication requirement on a case by case basis, you have a policy that might be beneficial. It could encourage more students to get their work out faster rather than sitting on it all until the end.

But if they take a hardline – no paper, no degree – lots of needless grief.


If publication is required on a case by case basis, rather than applying to all… Why make it a requirement, rather than simply encouraging publication?


personally-requiring publication is a political mine-field for candidates… the prospective policy is not a good plan.

Zen Faulkes

Having something listed as a requirement sends a much stronger signal than “encouragement.” It shows there’s a real expectation for something, and that the requirement will only be waived under extraordinary circumstances.

hapsci (@hapsci)

Nice write up of the conversation. My views are shared above..One thing I also brought up is that the publishing system is not perfect….

@AJEbrsary added – ‘My #scifi dreams: grad school as beginning of a multiclass RPG w/ branches 4 comms/research/management/pedagogy.’ Which I completely agree with.

PhDs should be about exploring, learning doing and making mistakes (and trying things that others cannot). If they are all about publishing papers how is a PhD different to a post-doc? Who will be free to try out the slightly more random and risky projects that may turn into something exciting?


Thanks for including Adrian’s comment, I couldn’t find a good spot for it in the post. Also, I was talking to a good friend of mine who graduated from my program a couple of years ago, and apparently a faculty member here once suggested a points system for graduation. For example, a first author paper in Science would give you a lot of points, enough to defend, BUT you could also accrue points through other activities, such as speaking at meetings, participating in outreach programs, getting teaching certifications etc. Seems like a genius plan to me, no idea why it never stuck….

And thanks for all your insightful comments on Twitter, too!


My PhD is not in science, but I feel that all PhDs have similarities, the main one being that the ASSESSMENT of your thesis is a peer reviewed process. I believe that it is beneficial to have articles peer reviewed during the candidature, but I know that when I submitted mine it really distracted me from my overall thesis and I’ve probably ended up taking longer to complete my PhD. My school encourages us to publish as much as possible and I think it’s ridiculous that a supervisor should discourage candidates from submitting to journals, because it is an important part of research. There’s nothing better than finding out that your work as successfully been peer-reviewed, and see it in print.

Michael Hiatt

Great write-up.

While I agree that issues surrounding the review and publication process are limiting and somewhat flawed, perhaps they are aspects of the training program of a PhD scientist that should be emphasized more. This leads back to the brief discussions we had yesterday regarding mentorship. What I like about a publication requirement, would be that it might shift the focus onto the PI/mentor to better facilitate and plan PhD student research to promote publishable science (per-review issues etc. aside, this remains the current model for disseminating science). I have been fortunate that each phase of my research was planned and conceived as both a thesis chapter and publication (even before I was mature enough as a grad student to think and plan that big a picture). Again, this brings us full circle to the larger issues with the publication and mentorship roles. In this light, perhaps it IS better to instead promote “publishable” science, but not require publication for defense or graduation.


Thanks for all your input the other day! It’s true, a significant problem with implementing a publication requirement is the publication process itself. But, like you said, that’s how science is currently disseminated. I have no real solution to that problem right now, although I am a fan of the post-publication peer review idea. Your comments about the quality of the mentor/mentee relationship are absolutely on the money. A friend of mine commented that you always know who will graduate first from a class because of who their PI is…

Richard Edwards

I think that someone should remind anyone proposing this that a PhD thesis *is* a peer-reviewed publication!

Different fields and different journals have different lag times. I published two papers from my PhD but neither came out before I submitted my thesis. Obviously, having your work peer-reviewed already (quite rightly) makes things easier – and makes you more marketable – but it should not be a requirement if the student and work performed is of the required calibre.


Also not a science PhD (I’m social science) plus I’m in Australia so this may have limited relevance as the way candidates are managed here is very different (much less structured and much more autonomous).

I have been able to get two papers published during the early part of my postgrad studies and I am now incredibly happy that I took the time to do that. In the latter part of my candidature my performance and progress has dropped because of health issues and because of my daughter’s health issues. Being able to point to my publications has meant that my PhD supervisors have evidence to support my capacity and can argue that the School (department) should continue to invest in me and my work and support me through this (hopefully) short-term difficulty.

If I didn’t have those papers under my belt I would look like any other PhD Mummy who can’t make it work and I would have probably lost the support of my School long ago.

However, I’m not a huge fan of compelling students to publish, for all the reasons that have been discussed here. Some work lends itself to publication, some does not. Disciplinary differences would make this requirement a considerable burden for some. I have published single author papers without too much pain but have worked on two multi-author papers that were horrendously difficult to get off the ground. If you are in a discipline that always does multi-author papers, you are having to put a key element of your PhD in someone else’s hands. No thanks.

I like the points idea that you mentioned. We have something similar to this for Honours graduates seeking a PhD scholarship. You get points for your Hons thesis, points for any publications, points for any relevant paid research work undertaken, etc. Everyone knows where they stand and has an opportunity to direct their efforts where it suits them.

Found your blog via Twitter, thanks to #phdchat 😀


Unfortunately, entire world of academia is completely dependent on scientific productivity measures mainly focusing on publication and citation records. In a current system anybody seeking a career in academia is simply required to present such records. Since grad schools are preparing young adepts of science to work and function in this environment, the publication requirement seems very reasonable and in the best interest of future PhDs. The problem is that in the contemporary world a vast majority of current grad students, and thus PhDs, will never pursue a career in academia, simply because of a lack of jobs for them. In this regard grad schools with their academia-only-fit curricula don’t prepare their students to face job market challenges. The publication requirement seems then to be a minor issue, at least from the perspective of entire system (but definitely not for individuals struggling with negative data).
On the other hand there’s also a problem of unpublishable, or negative, data. I believe that all research funded by public money should be available in public domain, doesn’t matter whether “publishable” or not. The lack of places where negative data could be published is not only a complication for many grad students, but a general big issue for whole scientific community.


Do you have to be the first author of the publication for it to qualify? Is this being implemented for all incoming students or retroactively?

David Stephens (@David_S_Bristol)

I am definitely not a fan of such a policy – especially where here in the UK one can get a PhD in 3 years. (Whether that is a good thing is another matter altogether!). Our department is very similar in size to yours – my impression is that most students do publish – either during or shortly after their thesis defence. Putting a requirement on it is only likely to increase the pressure to publish and in some (albeit rare) cases could lead to an increase in fraud (or at the least massaged data).

You also raise the issue of simply including published papers as thesis chapters. I am very definitely not a fan of this! Any graduate student leaving my lab leaves behind a comprehensive document detailing their methods, constructs, reagent sets, and almost all of their core data. That is the thesis. Binding a few reprints -which in many cases will have been written by the PI or by co-authors is just not the same. Thesis writing is an excellent exercise in self-discipline, scientific writing, and indeed manuscript writing (both primary research papers and reviews).

Overall – long may the thesis remain and I will certainly be wanting my students to produce a comprehensive thesis.


The 3-year British system has its pros ad cons for sure…I decided to come to the US for many reasons, including being able to take number of additional classes and sample 3 or 4 different labs before deciding where to do my thesis research. But at the end of the day I totally agree; having a publication requirement has the potential to cause an increase in fraud and a reduction in the ambitiousness of a student’s project.

As for including already published papers…I will actually be including two published and one (hopefully!) submitted paper in my thesis (as well as writing a ~10,000 word intro and an appendix covering my primary, failed, project). However, I am in the kind of lab where our PI expects us to write up our own work. I realize that this may not be the case in all labs, which is unfortunate, as having chapters written in advance spreads out the writing process.

David Stephens (@David_S_Bristol)

Thanks for replying Katy.
It is great that you have had such a productive PhD and that you are a) in an environment where you are encouraged to write up your own work and b) that you have the ability and aptitude for that. Not all graduate students do…..


Absolutely. And that’s what the publication requirement is supposed to help with. I just hope it works out for the benefit of the student, and that the advisers who don’t give their students such a supportive environment step up and change accordingly.

Anonymous! Safety first.

Hi, Katie PhD. Just came across this post and the dialogue on it. Very interesting and thoughtful!

Background: I’m in a very large molecular biology PhD program.

As somebody who is trying to graduate from a program that has recently (~3 years ago) implemented such a “rule,” my opinion is that it has gone badly. Like in the case of your program, the rule was put into place by a faculty member I know well with the intention of protecting students. However, the policy has only occasionally served to defend students from advisors trying to get rid of them; far more often, it has caused the retention of students desperate to graduate.

The idea was that the policy would cause advisors and students both to a) frame neater, tighter projects and b) focus on publishing smaller units instead of hanging on for years in order to get the high impact factor. Alas, no head honcho advisor is too keen on wasting data on small research articles when he can wait (he’s got time, right?) for the Big Paper. Project scopes are not changing, papers are not coming faster, and students are getting stuck for months and years in “end phase.”

Also (and I’ll try not to let my personal bitterness creep in here): student who were already dissertation phase when this rule was implemented (“grandfathered in”), are supposed to be exempt, but committees are largely not recognizing that reality. As somebody who is currently experiencing laboratory selection regret, I can confirm one of the ideas here: making publication a hard and fast rule prevents people from choosing the research they want, and instead encourages choosing the research that is likely to pan out/ be easier technically/ be “sexy science.”

Lastly, you are right about the non-academic career aspirers. As one of those, the publication requirement that is extending my already long PhD seems a waste of months I could be earning more than a pittance!

Interesting post. I hope this policy works in favor of the students in your graduate group!

Jon Ashley

Publishing papers before you submit the thesis is helpful in the fact that the work has already been peer reviewed. Therefore the examiners can’t really argue with the work.

Leave a comment


email (not published)